Software piracy is defined as the unauthorized use, duplication, distribution or sale of commercially available software. Although it has been argued that the existence of pirated software stimulates an interest in legal copies , estimated losses for the software industry over the past five years (19941998) are approaching $60 billion.
Losses due to software piracy are estimated by assuming that for each new personal computer sold there will be a set of accompanying software sales. For instance, to make the PC work the user must buy an operating system, and to make the PC useful the user would have to buy a set of productivity software, such as a word processor, and so on. On this basis, the shortfall between expected and actual sales must be due to software piracy, and the cost of these missing sales represents the financial loss to the software industry.
Using this method, the Software Publishers Association (SPA) and the Business Software Alliance (BSA) estimated that more than $12.3 billion was lost in worldwide sales in 1994, and $13.3 billion in 1995. Despite a drop to $11.2 billion in 1996attributed by SPA/BSA to cheaper software prices rather than a decline in piracy ratesthe figures for 1997 showed a rise to $11.4 billion.
In January 1999, the SPA merged with the Information Industry Association to form the Software and Information Industry Association (SIIA; www.siia.net). The latest SIIA report on software piracy, released in June 1999 , showed a continued decline in software piracy rates from its peak in 1995, but lost sales still amounted to $11 billion. Similar to the decline in 1996, however, the SIIA attributes this to the economic downturn in many of the highest offending countries, rather than a genuine decline in software piracy.
The main problem areas include Eastern Europe, Latin America, and the Asia-Pacific region, while the percentage of copies in use that are pirated are estimated to exceed 90% in countries such as Russia, China, Indonesia, Vietnam, Lebanon and Oman. Even though the U.S. has one of the lowest piracy rates in the world (25%), the size of the software market has meant that, even at current levels of piracy, the U.S. accounted for over $2.8 billion of lost software sales in 1998.
Clearly, software piracy is a major worldwide issue, and as the millennium brings a more software-intensive business environment, reasons for and measures to prevent software piracy will remain a multibillion-dollar industry.
What Can Be Done
The SIIA suggests a three-pronged attack on software piracy using legislation, effective enforcement, and education of the public. This approach is based on the conviction that buying and selling pirated software is essentially trafficking in counterfeit goods. Indeed, most countries provide legal protection for software by extending copyright, patent, contract and trade secret legislation, and by recognizing software as another type of literary and artistic work subject to intellectual property right (IPR) protection. The World Intellectual Property Organization (WIPO) Copyright Treaty governs international protection for intellectual property rights under the auspices of the World Trade Organization.
Given that software piracy is a criminal offense it would seem the act of making, distributing, or buying pirated software is simply a question of the lack of morality of the pirate versus the non-pirate , or due to a lack of understanding of the copyright laws . On this basis, the SIIA's calls for increased legislation and enforcement would appear correct.
However, there is also evidence to suggest that another factor strongly motivates software piracy. Alongside the lack of censure for piracy and the low likelihood of being caught, the most common reason offered for pirating software is the high cost of legal software . This would suggest a missing element in the SIIA's fight against software piracy, that is, a review of the cost of legal software. Indeed, it is interesting to note that in the two instances since 1994 when the global losses due to software piracy declined (1996 and 1998), the SIIA explains the decline simply in terms of cost.
The extent to which either legislation, enforcement, or cost would be effective in deterring the buying of pirated software can be tested by analyzing the responses of subjects to a number of ethical scenarios in which censure, availability, and cost are issues. When presented with an ethical scenario, however, subjects tend to be sensitive to context . In other words, an ethical judgment toward an act such as software piracy is based not simply on the legality or otherwise of the act, but on the context in which the act takes place.
For instance, the unauthorized copying of an expensive system may be seen as unacceptable in the context of stealing from one's own company, but may be seen as acceptable if it is justified in the context of a genuine need . These situational effects have been found to differentiate the ethical responses of males and females  and across organizations , suggesting gender and culture has an effect on the way ethical scenarios are interpreted.
In order to understand the context under which software piracy flourishes, it would seem appropriate to analyze a sample population of those people most likely to be engaged in the behavior. Since young, male students are seen as the archetypal software pirates [3, 9], a survey of students would seem to be the most appropriate target sample.
Furthermore, in order to elicit true responses to familiar situationsrather than weak responses to hypothetical or unfamiliar situations it would seem appropriate to conduct the survey in a location where software piracy is a well-known phenomenon. Hong Kong remains on the U.S. Special 301 review watch list for unfair international trade practices and is often cited by the SIIA for the level of software piracy. As such, the respondents in this study are students in Hong Kong.
Hong Kong is famousor from SIIA's perspective, notoriousfor the existence of a number of well-known arcades that openly sell pirated copies of software. Probably the best known of these is the Golden Arcade in Sham Shui Po on the Kowloon peninsula. With the onset of CD-ROMs, pirated software is now almost exclusively sold as CD sets. The CDs are typically sold in packs of four for HK$100 (approximately $12), with each disc containing complete software suites such as MS Office, or sets of applications ranging from Web browsers to programming languages. Continued crackdowns by Customs officials have seen the CD pirates move more cautiously; CDs are no longer openly sold in the Golden Arcade, but nearby alleys and back rooms continue the trade.
Intellectual property rights are protected in Hong Kong under a number of laws that have their origin in the U.K.'s Copyright Act (1956). In particular, the Copyright (Computer Software) Amendment Act (1985) was extended to Hong Kong in 1988, and establishes computer software as a literary work and thus protected by copyright legislation. An amendment in 1995 set the penalty for software piracy to a maximum fine of HK$50,000 per copy (approximately $6,400), and four years imprisonment. A further amendment in 1997 allow Customs officials to confiscate CDs simply on the basis they are suspected to be fake, thereby transferring the burden of proof of authenticity from the Customs Officials to the owner.
Even with this set of legislation in place, software piracy remains a problem in Hong Kong. In October 1997, for instance, Microsoft experts examined over 3,000 CDs seized during 24 raids by Hong Kong Customs officials on CD-ROM dealers. The CDs contained business and game programs with an estimated value of over $900,000. Two people were prosecuted; one received a four-month prison sentence, while the other was fined $26,873. In common with many of the Asia-Pacific countries, the software piracy rate remains high in Hong Kong, with an estimated software piracy rate of 59% in 1998, which amounts to an estimated $88.6 million in lost sales.
In order to launch an effective campaign against software piracy it is important to identify those factors typically seen as motivating software piracy. Following the literature, it is suggested that three features characterize software piracy in Hong Kong: censure, availability, and cost.
Censure refers to the perceived threat of punishment that accompanies the buying of pirated software. The illegality of software piracy is the focus of the suggestion that to buy or not to buy pirated software is an ethical judgment. Indeed, SIIA's principal means of attacking the software pirate is to increase the actual and perceived censure associated with the act, through intellectual property rights legislation, adequate enforcement of those laws, and education of the public about the nature of intellectual property rights.
Availability refers to the perceived ease with which pirated software can be bought. The suggestion here is if pirated software is readily available and easy to purchase, then the likelihood of purchasing pirated software increases. Potential customers need to know about a product before they can buy. Furthermore, if an illegal product is widely available there can be doubt in the mind of the purchaser that they are doing something illegal, and so, the apparent ethical dilemma is somewhat diluted.
Cost refers to the perception that software vendors are charging too much for the software. The suggestion here is if it is believed the price of software is beyond the purchasing power of those who would like to use it, pirated software provides an alternativealbeit illegalmeans of gaining access to the software. Driven by a need to use the software, the software pirate may be willing to risk censure and seek out pirated software in order to satisfy the need.
In Hong Kong, there is a low perception of censure for software piracy, high availability, and high cost of legal software. In particular, there have been few prosecutions of software pirates in the Hong Kong press, while the existence of a number of arcades throughout Hong Kong makes it clear to most people that pirated software is readily available and at a price considerably lower than for legal copies. For brevity, this situation is coded as L-CENSURE, H-AVAIL and H-COST.
An open question is whether changing the context under which software piracy exists would actually have any impact on the buying behavior of those involved. In other words, if a lack of censure is a key reason for people, would people continue to buy pirated software if the threat of censure was high? This question can be answered by reversing the three contexts noted previously. The reversed situation is there is high censure for buying pirated software (H-CENSURE), low availability (L-AVAIL), and little difference in cost between legal and pirated software (L-COST).
This reversal would indicate the likely effectiveness of campaigns directed at these issues. Of course, by asking the respondents to hypothesize about their actions in a context that does not exist, there is the danger the responses will be less insightful, simply because they may not be able to clearly envisage what their behavior would be.
These six conditions are framed using direct statements about the level of censure, availability, and cost and the respondents' likelihood of buying pirated software under those conditions. The statements are presented in Table 1. Although the statements relate to likelihood of buying, intentional statements are often used as predictors of likely behavior.
The questions appeared on the survey form with a seven-point Likert scale, ranging from 1=Strongly Agree to 7=Strongly Disagree, with 4=Neither as the mid-point. The survey was entitled "A Survey of the Reasons for Buying Pirated (illegal) Software," and as shown in Table 1, the statements make it clear the software being bought was pirated. Each of these points is designed to attune the respondent to the fact that buying buying pirated software is illegal.
The survey involved students taking information system courses in the Faculty of Business at the City University of Hong Kong. Clearly, an element of trust is required before outlining their engagement in an illegal activity. As such, the students were informed the survey was about software piracy and participation was voluntary. A total of 243 usable responses were received, of which 122 were female and 121 were male.
As shown in Figure 1, 81% of the respondents report they buy pirated software on a regular basis, with a sizable minority (29%) buying every month, and 3% even reporting they buy several times a week. The most popular pirated software bought was spreadsheets, followed by programming languages, databases, word processors, and statistical packages. Other software mentioned included email, graphics, and game software. Only 7% claim to have never bought pirated software. This set of respondents clearly know about software piracy and are therefore likely to provide a true insight into what motivates those people who buy pirated software.
The idea the typical software pirate is a young male is confirmed by comparing buying rate to gender and age. Applying a Mann-Whitney U Test with gender as the grouping variable gives U=5372.5 (Z=-3.6660, p=0.0002). Applying Kruskal-Wallis ANOVA by ranks with age as the grouping variable gives H=24.69 (p=0.0000). The age result must be viewed with some caution since more than half the sample fell in the 2125 years range, thus biasing the result toward this single category.
The expected versus actual responses are classified in Table 2. Since low censure (L-CENSURE), high availability (H-AVAIL) and high cost (H-COST) are expected to be reasons for buying pirated software, the respondents should agree with these context statements. This is classified in Table 2 as a Y in the Expected column. Since H-CENSURE, L-AVAIL and L-COST are meant to show the effect on software piracy of reversing these contexts, it was expected the respondents would not agree with these context statements. This is classified in Table 2 as an N in the Expected column.
The actual response of the sample is classified in Table 2 using the Median score. If the median is less than 4, then the overall response is taken to be they agree with the context statement. This is classified in Table 2 as a Y in the Actual column. If the median is 4, then we have an undecided response equivalent to a "Neither" response on the Likert scale. This is classified in Table 2 as a U in the Actual column. Finally, if the median is greater than 4, then the overall response is taken to be that they would not agree with the context statement. This is classified in Table 2 as a N in the Actual column.
Using this scheme, Table 2 summarizes the responses to the six context statements, with strong agreement to the first three context statements in evidence. However, the effect of reversing the context brings a mixed response.
The indecision of the respondents when censure becomes an issue (H-CENSURE) is notable. Although a Wilcoxon matched pairs test shows a significant difference between the L-CENSURE and H-CENSURE responses (Z=10.6643, p=0.0000), a median score of 4 suggests the respondents are undecided about buying pirated software when censure is high. The indecision may be based on defining the level of censure involved. While the H-CENSURE context statement mentions a "fear of punishment," the level of punishment itself is left undefined.
The disagreement among respondents, therefore, is probably based on assessing what the censure would actually entail as well as the degree to which they would fear any kind of punishment. In short, given the actual L-CENSURE context in Hong Kong, a further shift in responses would require a clearer definition of the punishment involved. This result fits with SIIA's goal of increasing the actual and perceived censure associated with software piracy through both legislation and an education of the public about the issues involved.
At first glance, availability does not appear to be an issue since the respondents broadly agreed they would buy pirated software under both contexts (H-AVAIL and L-AVAIL). Although a Wilcoxon matched pairs test shows a significant difference between the H-AVAIL and L-AVAIL responses (Z=8.8605, p=0.0000), even when pirated software is not freely available the median score suggests that respondents would continue to buy.
This general agreement appears to be based on gender, however, since 18 of the 19 respondents that strongly agreed they would continue to buy were male. This would suggest a more focused approach is needed in the battle against piracy, with a campaign to reduce the macho image of buying pirated software.
The only context that met the expectation of switching from agreement to disagreement when the context is reversed was the context of software cost. A Wilcoxon matched pairs test, not surprisingly, shows a significant difference between the H-COST and L-COST responses (Z=12.4184, p=0.0000), but more importantly, the median score for the L-COST context suggests respondents would not continue to buy pirated software if the cost of legal software is reduced.
The issue of the price being charged by software vendors is not a part of SIIA's attack on software piracy, although it clearly has a significant impact on the perception of the respondents in this survey. In this context, legislation and consumer education are not the key factors in the buying of pirated software in Hong Kong, but simply the perception that software vendors are charging too much for their products.
The level of software piracy is high among the 243 respondents, with 81% of the respondents buying pirated software on a regular basis. The study confirms the typical view of the software pirate as a male, since there is a significant difference between males and females in the frequency with which they buy pirated software.
There is strong agreement from the respondents in terms of buying pirated software in the context of high availability of pirated software, low censure for buying, and the high cost of legal software. Reversing the contexts (such as low availability, high censure, and low cost) resulted in a significant shift in their responses, suggesting the potential to reduce software piracy by addressing these three issues. However, only the cost issue showed the expected switch from agreement to disagreement.
Motivated by the belief that if consumers can afford to pay for the hardware they should be expected to pay for the software, the SIIA focuses more on the use of legislation and education to enforce intellectual property rights. While the SIIA may be correct to note the economics involved in software piracy they may, in fact, have identified a prime motivating factor for piracy: That of the cost of software products relative to the cost of hardware.
For instance, in the fall of 1998 the cost of a personal computer with a Pentium-II 450MHz processor, 10GB hard drive, and 128MB of RAM from online retailers such as Dell (www.dell.com) and Compaq (www.compaq.com), was around $2,700. By the Fall of 1999 the price of a similarly configured personal computer had declined to around $1,500.
In the same time frame, however, software consumers will have been faced with upgrading their operating system and application suites, with further releases already being previewed in the computer literature. While upgrades are not compulsory, advertising campaigns and the addition of Web-related functionality would convince most software consumers that an upgrade is worthwhile if it can be afforded.
In short, software piracy appears to be due in part to a perception by consumers that software has not followed the same path as hardware and clearly demonstrated an increase in value for money. The view from Hong Kong would seem to be that the cost of remaining up-to-date underpins a flourishing market for pirated software.
©2000 ACM 0002-0782/00/1200 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2000 ACM, Inc.